A new non-recurrent congestion detection and traffic re-routing model with ant consensus behaviour in house hunting algorithm

The increasing traffic congestion has led to the development of various traffic routing strategies to improve the network's performance. However, most of these strategies have focused on recurrent congestion (RC), which refers to the regular and predictable congestion during peak hours. On the other hand, non-recurrent congestion (NRC) is unpredictable and occurs due to accidents, road closures, or other unforeseen events. NRC affects road users and traffic monitoring authorities in various ways. Hence, it is essential to address this issue with practical solutions to alleviate its impact on transportation systems and communities. The challenge in solving NRC, which involves distinguishing between RC and NRC, is further compounded by the need to establish clear standards for usual and unusual traffic indicator measurements, which are essential for accurate identification. NRC must be timely and precisely identified to enable prompt management to minimize the wastage of resources and the potential for another NRC event. Various approaches aimed to alleviate congestion by improving traffic flow, increasing capacity, and optimising traffic demand have been proposed. However, existing solutions often prioritise individual drivers' convenience and minimise travel disruptions without considering the overall network performance, which is crucial during NRC. To be effective, NRC management needs near real-time traffic monitoring and strategies for re-routing.